Random and Stochastic Structural-Acoustic Analysis

Real engineering systems include some levels of uncertainty that cannot be captured by the available deterministic analysis methods. The modeling of such systems involves diverse parameters already characterized by some simplifications to the real physical behavior. This chapter addresses the methods for uncertainty quantification in structural-acoustic analysis. The major focus is on the spectral non-sampling methods and stochastic finite element method (SFEM) modeling of vibroacoustic problems. The methods available for discretization of random parameters have been presented. Among them, the Karhunen-Loève and polynomial chaos expansions are discussed in detail. The SFEM in combination with these expansions is formulated for general structural vibration and vibroacoustic problems. Two classes of numerical SFEM procedures, intrusive and nonintrusive methods, are reviewed. Numerical examples show the application of methods for practical structural-acoustic problems, where experimental modal data are used to identify the uncertainty in material parameters.